Redefining Virtual Instruments

The NEATBrain engine uses a hybrid synthesis technique to accurately model real instruments.

First, the Reassigned Bandwidth-Enhanced Additive Sound Model is used to extract the building blocks of the original signal. These are repitched across the keyrange without artifacts, then truncated to form digital waveguides.

These waveguides are loaded into the “Memory” (the actual playable instrument), and looped with a single-cycle sampler. Finally, the Residue (noisy signal) is played alongside these waveguides to accurately reconstruct the original sound. This technique greatly reduces memory usage while still allowing features found in traditional Sample Libraries, such as Round Robin.

Because the primary elements are separated, the end-user can sculpt the sound in a variety of ways, including adding and adjusting Tone Profiles, as well as modifying the “expressiveness” of the instrument through the use of various Amplitude and Pitch Modulators.